Xenia assisted reciprocal recurrent selection in maize
--Wolfe, RI

Reciprocal recurrent selection (RRS), a breeding system in which two populations undergo repeated cycles of selection based on performance of hybrids between them, has been shown to be highly effective for improvement of a number of characteristics, including yield. However, the necessary hand crossing, record keeping, and the time generally needed to complete a full RRS cycle, limit the power of the method.

Maize, being naturally open pollinated, with the male and female portions of the plant separated, is easier than most species to cross by hand. Nevertheless, if one could retain the selective power of RRS in maize while eliminating most of the hand crossing and record keeping, and if one could shorten the RRS cycle time to one year, it should be possible to substantially improve efficiency.

This may be possible, using xenia to identify pollen source. A possible term for such a breeding method is, Xenia Assisted Reciprocal Recurrent Selection (XARRS).

There are two major forms of visible xenia in maize, aleurone color and type of endosperm. With the best choice from the species of a genetic marker from each type, it may be possible to follow pollen movement from each of two RRS populations, while maintaining effective population identity or integrity.

To make XARRS work, two populations must be set up, one homozygous, or nearly so, for a dominant aleurone color, and homozygous for a recessive shrunken endosperm. Pollen from this population must produce seed identifiable by color on the second population. This second population would be homozygous for a recessive counterpart of the colored allele in the first population, and homozygous, or nearly so, for the dominant starch-producing allele. Pollen from this population must produce normal starchy seeds on the first population.

The ideal for XARRS would be to have only one major gene of each xenia type in the breeding material under RRS, each gene with only two alleles, and all four seed types easily distinguishable. That is, there would be a color gene, with two alleles such as yellow and white, and a gene conditioning seed filling, such as normal starch and sugary, su. Even if one could get such, it would probably prove to be important to select strongly against any modifiers that appear in the two populations.

These two populations would be grown together and allowed to intercross. Some of the resulting inter-population hybrid seed, recognizable as double dominants, would be grown out and tested. The rest of the seed from each plant would be kept in reserve. Information gained from testing would be used to rank the plants from which the seed was obtained. Intra-population seed from the best of the reserve plants would be used to form the next crossing block. If the crossing block were to be grown at a suitable winter increase site, one XARRS cycle could be completed per year, without any hand crossing in this part of the breeding program.

If the two marker genes, with two contrasting alleles each, function well, it would not be necessary to set the populations up with absolute, or even extremely high levels of homozygosity of the two dominant markers. Some consequent mixing between the two populations would be good for their evolution. However, if mixing produces complicated segregations of seed color, more attention would have to be paid to starting the two populations out highly homozygous for their respective dominant markers. To maintain or improve the level of homozygosity, seed of plants harvested from the crossing block having any double recessive seeds would be discarded.

Several XARRS cycles should build up considerable improvement in performance of inter-population hybrids. The world's best germplasm could be incorporated into both populations on a continuing basis.

A small back-crossing program, using the best material as recurrent parents, would be run to eliminate markers that are unacceptable for the intended use of resulting hybrid cultivars. With this being done where considered necessary, elite material from each population would be selfed to produce inbreds. Hybrids for commercial release would then be developed between the inbreds of the two populations, using standard maize breeding procedures. At some point, hopefully sooner rather than later, the best of such hybrids should begin to outperform the best commercial hybrids available from other sources.

Within an XARRS breeding program, a possible set of designations for seed types and the two populations would be:

-"Dash" for Dark-shrunken seeds, and the population designated by them

-"liSt" for light-Starchy seeds, and the population designated by them

-"DkSt" for the Dark-Starchy inter-population hybrid seed

Over the past three years I have been working on setting up genetic material for XARRS in barley, where two genetic male sterile genes, one in each of the two populations, are necessary to force out-crossing.

During this period I have also been trying to demonstrate the method in maize, where genetic male sterility is not needed. Our environment at Lacombe is not suited for commercial production of field corn. However, in sheltered locations over the past several years it has been possible to get some viable seed on the earliest maturing plants.

In 1996 I grew out some blue-starchy seeds obtained from yellow sweet corn cobs, mainly of the local early maturing cultivar Sunnyvee. These seeds had been produced from pollen of other maize plants growing nearby. I had been planning to use yellow and white as the color markers, but found that the color contrast sometimes virtually vanished. Upon seeing the rather strong xenia effects on sweet corn of pollen, carrying a gene for a dark blue aleurone color, I decided that blue seed might contrast better with yellow than the contrast of the yellow-white combination.

This past year, 1997, I planted some of the blue-shrunken seeds from 1996 in rows adjacent to a number of yellow starchy lines, including LX511, an early maturing hybrid from Dr. Malcolm MacDonald, Lethbridge, Alberta, and some material received from Dr. Bob Hamilton at Ottawa. These were designated, `Ble d'Inde Gaspe' flint and sugar, `Brandon Research Centre Synthetic', and `V360 Spanish Flint'. The Gaspe material was the earliest of any in the crossing block.

Cobs of plants from both the yellow-starchy and blue-shrunken seed planted in 1997 had some blue-starchy kernels. These were the result of inter-crossing of the two seed types. Some of these blue-starchy seeds are being grown out in Chile this winter, 1997-1998, and are to be selfed. It will be interesting to see how many, if any, of the resulting plants have a definite one-gene segregation for color.

The segregation of seed color on the plants from the blue-shrunken seed in 1997 was quite variable. It was obvious that there was more than one gene influencing color. Consequently, it is possible that the white-yellow color combination will work better. If this latter color pair is to be used, selection of the palest �white� white-starchy and the most strongly yellow yellow-shrunken seed for the XARRS crossing block might keep the color modifiers at bay.

Return to the MNL 72 On-Line Index
Return to the Maize Newsletter Index
Return to the MaizeGDB Homepage